inference-optimization/NemotronH-0.3B-A0.3B
Different architecture vs nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-Base-BF16. Different architecture: only 0 of 51023/168 tensors match by name and shape.
Different architecture vs nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-Base-BF16. Different architecture: only 0 of 51023/168 tensors match by name and shape.